Predictive Model for Carbapenem Resistant Pseudomonas aeruginosa on Admission to Acute Care Hospitals Public

Abstract

Background: Carbapenem-resistant Pseudomonas aeruginosa (CRPA) is a multidrug-resistant bacteria that frequently causes healthcare-associated infections (HAI). Since CRPA cannot be treated with carbapenems, treatment options are limited. CRPA can colonize patients and the healthcare environment, increasing the risk of healthcare-associated transmission and outbreaks. We aim to evaluate risk factors for CRPA carriage for all patients hospitalized at an academic healthcare system in Georgia.

Methods: We performed a case-control study using electronic medical record (EMR) data from 4 healthcare facilities within a single healthcare system from 1/1/2014 to 12/31/2021. Cases were defined as adult patients (>18 years old) with their first inpatient encounter where a clinical culture identified P. aeruginosa resistant to meropenem or imipenem collected on the day prior to admission or within the first 2 hospital days. Only the first CRPA culture from patients was included. Controls included admissions from the same hospital during the same month and year that the case sample was collected. the dataset was split into training and testing datasets. Univariable logistic regression was used to determine crude associations between the model covariates and the outcome. Best subset selection was used to select the final model. A multivariate logistic regression model and ten-fold cross-validation were used to estimate coefficients for predictors of interest in the model. The models were assessed by receiver operating characteristic (ROC) curves and area under the curve (AUC). A secondary analysis was conducted to exclude patients with cystic fibrosis using the same methods as the primary model.

 

Results: We identified 521 cases and 560,623 controls. The AUC of the primary model was 0.80. The AUC of the secondary model was 0.72. Cystic fibrosis was the most significant predictor of CRPA. Diabetes as well as days of carbapenem treatment, infection diagnosis, and ventilation in the last year were significant predictors of CRPA carriage independent of cystic fibrosis.

 

Conclusion: We developed a model with good predictive ability. Implementation of this model in healthcare facilities could help with earlier identification of patients with CRPA and decrease the risk of healthcare-associated transmission. Future studies could prospectively validate this model’s performance to accurately identify patients with CRPA carriage on admission.

Table of Contents

Background and Public Health Relevance.......................................................1

Methods......................................................................................................2

Results........................................................................................................4

Discussion...................................................................................................6

Conclusion..................................................................................................9

Table 1: Patient Characteristics and Univariable Analysis ..............................10

Table 2 Primary Multivariable Model............................................................11

Figure 1 ROC Curve for Primary Multivariable Model.....................................11

Table 3 Results for Secondary Analysis.........................................................12

Figure 2 ROC Curve for Secondary Analysis ..................................................12

Reference……………………………………………………………………………...……...13

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Epidemiology
Subfield / Discipline	Epidemiology - MPH & MSPH
Degree	M.P.H.
Submission	Master's Thesis
Language	English
Research Field	Health Sciences, Epidemiology
Mot-clé	epidemiology predictive modeling healthcare epidemiology Pseudomonas aeruginosa infection prevention CRPA
Committee Chair / Thesis Advisor	Fridkin, Scott, Emory University
Committee Members	Howard-Anderson, Jessica, Emory University
Partnering Agencies	Emory University schools, faculty or affiliated programs

Dernière modification

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Predictive Model for Carbapenem Resistant Pseudomonas aeruginosa on Admission to Acute Care Hospitals ()	2024-04-25 19:35:21 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions